SAFE
The core of emergency medicine is physically laying hands on undifferentiated, often uncooperative patients: airway management, central lines, chest tubes, reductions, ultrasound at the bedside, and deciding in minutes who is dying and who can wait. AI is genuinely good at the paperwork layer — ambient scribing of ED notes, discharge instructions, coding, ECG and imaging pre-reads, triage risk scores — and that will compress documentation time rather than the physician's role. Licensure, EMTALA obligations, and personal malpractice exposure mean a board-certified human owns every disposition decision.
Most of this decline happened after 2021 — it is not the pandemic dip.
This line is counted by the Bureau of Labor Statistics — the one figure on this page that isn't a judgement of ours. Headcount moves on demand, offshoring, demographics and the business cycle, and automation is one term among several, often not the loudest.
So a falling line is not evidence that AI did it, and a rising one is not evidence that it won't. Both happen in this register: some occupations resist automation and shrink anyway, others are highly automatable and keep growing.
BLS projection, 2024–2034
+2.7% 36,100 → 37,100 on the projections basis
Hard to automate, and growing
The work resists current AI and the BLS projects +2.7% more of these jobs by 2034. Note that safe does not mean well paid — several of the fastest-growing resistant occupations are among the lowest paid on the register.
Different clocks. The score is what current AI could do to this work today. The projection is how many of these jobs will exist in 2034. Everything between the two — how fast employers actually adopt, whether demand grows in the meantime — is why they can point opposite ways without either being wrong.
~1,000 openings a year on average, including replacing people who leave.
PhysicianIntensivistTrauma DoctorEmergency DoctorAttending PhysicianEmergency PhysicianMD (Medical Doctor)Wound Care PhysicianUrgent Care PhysicianCritical Care PhysicianCritical Care IntensivistDisaster Medicine PhysicianAttending Emergency PhysicianEmergency Medicine SpecialistER Doctor (Emergency Room Doctor)Pulmonary Critical Care PhysicianCritical Care Intensivist PhysicianPediatric Emergency Medicine PhysicianEmergency Department Doctor (ED Doctor)ER Physician (Emergency Room Physician)Emergency MD (Emergency Medicine Doctor)Emergency Medicine Physician (EM Physician)Emergency Department Physician (ED Physician)
Holding it up: liability shield . Weakest point: trust premium .
Tasks largely resist digitisation Documentation, coding, ECG pre-reads and discharge-instruction generation are already being eaten by ambient scribes and algorithmic triage scores, which is why this sits at 15 rather than 19 — but intubating a bleeding trauma patient, reducing a dislocated shoulder, or placing a chest tube in a hypotensive pneumothorax has no digital substitute.
Hands-on in uncontrolled environments An 18 reflects that essentially every shift involves procedures on unstable, sometimes combative or contaminated patients in a resuscitation bay you don't control — blood, vomit, C-collars, code carts — with the remaining 2 points off only because charting and consult calls happen at a workstation.
Licensed human required and personally liable State medical licensure plus board certification is non-negotiable, EMTALA makes you personally responsible for screening and stabilizing every patient who walks in regardless of ability to pay, and your name is on the disposition when a missed dissection becomes a malpractice claim — there is no higher configuration than this.
Exists to be accountable for ambiguous calls Deciding in under ten minutes whether an undifferentiated abdominal pain goes to the OR, the CT scanner, or home — and running a code where you call time of death — is the archetype of an ambiguous high-stakes call made on incomplete data with no one senior to defer to at 3 a.m.
The verdict above describes this occupation as a whole. Almost nobody does the typical version of a job — tick what's actually in your week and see how your own mix sits.
Your task mix speaks to task resistance (15/20 here) — how much of the day's work current AI already does. That is the dimension the boxes above are about.
It cannot move the other three. Liability shield (20/20) is whether the law requires a licensed human to sign. Trust premium (15/20) is whether buyers specifically pay for a person. Judgment and accountability (20/20) is whether the role exists to own consequential calls. Those are facts about the occupation's standing, not about which tasks are in your week — a paralegal who does only trial exhibits still holds no licence. Together they are 55 of this occupation's 88 points (63%).
Embodiment (18/20) is also a property of the work rather than the worker, but we don't tag individual tasks as physical or not, so the picker can't tell you anything about it. That's a limit of this tool, not a claim.
Did we get the list right? Tell us what's missing — the tasks are written from the outside, and you're reading this from the inside.
The moves above are yours to make. This is the other half: what would have to change in the world for the occupation itself to score higher. None of it is in any one person's gift, but it is where the floor actually comes from. Scores here are not a one-way ratchet. Only two of the five dimensions — task resistance and embodiment — track what machines can do. The other three track law, what buyers will pay for, and who is answerable, and those move in both directions, often in response to the same pressure AI creates. If every lever below landed, this occupation would score around 93/100, still SAFE.
Task-mix shift: ambient scribes (Nuance DAX, Abridge in ED settings), autonomous ECG and CT pre-reads, and auto-generated discharge instructions strip the documentation/coding tier out of the shift, leaving a residual day that is almost entirely procedures, resuscitation, and undifferentiated-complaint disposition. The routine tier shrinking makes the measured remainder less automatable, not more.
If ED boarding and admitted-patient holds continue to worsen (ACEP's boarding crisis declarations, post-2022 data), the physician's day shifts further toward hands-on management of multiple unstable held patients plus procedural sedation, reductions, and difficult airways — physical work in an environment that is unpredictable by definition.
A narrow route only: if CMS or a state licensing board requires that the ED attending personally perform a documented in-person evaluation for high-acuity ESI-1/2 presentations rather than accepting APP-with-remote-supervision models, the human-attending encounter becomes a purchased requirement rather than a staffing preference. Watch state scope-of-practice fights and ACEP's position statements on independent APP practice in the ED.
The limit. liability_shield and judgment_accountability are already at 20; EMTALA plus board certification plus personal malpractice exposure cannot be tightened further in a way the register would register. At 88 this occupation is essentially at the practical ceiling — the realistic movement is a few points of task-mix consolidation, not a structural change.
| New York-Newark-Jersey City, NY-NJ | 3,070 | $289,440 -14% |
| Chicago-Naperville-Elgin, IL-IN | 960 | $239,200 -29% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 770 | $239,200 -29% |
| Boston-Cambridge-Newton, MA-NH | 620 | $356,230 +6% |
| Minneapolis-St. Paul-Bloomington, MN-WI | 550 | $367,120 +9% |
| Detroit-Warren-Dearborn, MI | 520 | $447,750 +33% |
| Milwaukee-Waukesha, WI | 500 | $239,200 -29% |
| Indianapolis-Carmel-Greenwood, IN | 490 | $284,930 -15% |
| Huntington-Ashland, WV-KY-OH | 40 | $531,340 +58% |
| Providence-Warwick, RI-MA | 300 | $476,890 +42% |
| Anchorage, AK | 70 | $472,140 +41% |
We have no reported case of a named organisation automating this occupation. Not one deployment, not one announcement.
That is worth saying out loud next to a score of 88. The verdict above is about what the work exposes — what current AI could do to these tasks. It is not a claim that anyone has done it. For this occupation those two things have come apart completely: the capability argument is on this page, and the evidence column is empty.
Has AI actually changed your work? One tap, anonymous, and the running tally is public. Nothing else is asked of you.
Rather than check back: get the digest and we'll tell you what changed — or watch a single occupation from its own page.